Does Ziva generate models the way text-to-3D tools do?

No. Ziva does not use text-to-3D or image-to-3D neural models, including Rodin, Meshy, Tripo and Hunyuan 3D. Its agent writes Blender Python that combines shapes, materials and modifiers, so the output is editable low-poly geometry.

Which plans and platforms support Ziva's 3D Model Generator?

It is limited to the Pro and Ultra plans; Hobby and Basic users see an upgrade card. It runs on macOS with Apple silicon, Windows x64 or ARM64, and Linux x64, but not on Intel Macs or Linux ARM64.

What can't Ziva do with these models?

It has no built-in rigging, animation, retopology or LOD generation, and reference images only guide the script rather than being reconstructed into a matching mesh.

Image GenerationAI Coding AssistantsZenn AI/ML3m agoNov 8, 2026, 06:00 JST

Ziva guide: 296-triangle chest built inside Godot

Send on LINE

3 Key Points

What happened
Ziva.sh published a walkthrough of its in-editor agent Ziva, which on Pro or Ultra plans turns a plain-language prompt into a saved .blend model, producing a 296-triangle chest and a 70-triangle pine tree on Godot 4.7.2.
Why it matters
Because the model is built from Blender scripts rather than a text-to-3D neural generator, developers can ask for proportion or color changes in the same chat, though the approach suits simple low-poly props rather than organic characters.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Get these free every morning

Context & Analysis

The guide draws a clear line between two ways of getting 3D assets into a game engine. Cloud generators return a finished mesh and stop there, while Ziva's approach keeps the model as a script-defined .blend file, which is why the same chat can later be asked to enlarge a lid or darken a tree. That editability comes at a cost in scope: the two examples in the test were deliberately simple, a chest with three materials and a tree with two, both using flat colors and no textures.

The walkthrough is also candid about setup friction. The 3D Model Generator is gated behind the Pro and Ultra plans and needs a one-time Blender download, which the author measured at 365.5 MB on Linux. On Godot 4.3 and later, Ziva configures its own Blender path automatically if the editor setting is empty, which removes one manual step that .blend imports normally require.

For everything beyond simple props, the guide points readers elsewhere: characters and organic shapes to hosted generators, free offline image-to-3D to TRELLIS.2, and ready-made asset sets to libraries like Kenney.

FAQ

Does Ziva generate models the way text-to-3D tools do?: No. Ziva does not use text-to-3D or image-to-3D neural models, including Rodin, Meshy, Tripo and Hunyuan 3D. Its agent writes Blender Python that combines shapes, materials and modifiers, so the output is editable low-poly geometry.
Which plans and platforms support Ziva's 3D Model Generator?: It is limited to the Pro and Ultra plans; Hobby and Basic users see an upgrade card. It runs on macOS with Apple silicon, Windows x64 or ARM64, and Linux x64, but not on Intel Macs or Linux ARM64.
What can't Ziva do with these models?: It has no built-in rigging, animation, retopology or LOD generation, and reference images only guide the script rather than being reconstructed into a matching mesh.

Zenn AI/MLRead Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.